MicroRNA-210 as a Novel Therapy for Treatment of Ischemic Heart Disease

293FT cells (Invitrogen) were used to generate recombinant replication-deficient lentivirus used for in vitro assays as described.⁸ Mouse HL-1 cardiomyocytes were cultured in Claycomb media supplement 10% fetal bovine serum (Sigma), 0.1 mol/L norepinephrine (Sigma), 2 mmol/L l-glutamine (Invitrogen), and penicillin/streptomycin (Invitrogen) in a humidified 5% CO₂ incubator at 37°C. Hypoxia was achieved by placing cells in a hypoxia chamber filled with 5% CO₂, 1% O₂, and 94% N₂ at 37°C. At different time points during hypoxic treatment, cells were harvested for analysis of miR-210 levels. Angiogenesis factor antibody array (Panomics Inc, Fremont, Calif) was used to investigate the angiogenic potential of miR-210. Apoptosis assays were performed after 48 hours of transduction with lentivirus carrying miR-210 precursor (Pre-210), lentivirus carrying miR-scramble (Pre-Scr), lentivirus carrying antisense of miR-210 (Anti-210), and lentivirus carrying antisense of miR-scramble (Anti-Scr) using a Caspase-Glo 3/7 Assay (Promega) according to the manufacturer’s instructions. The caspase activities for all samples were normalized to that of an equal protein amount. The data obtained are from experiments performed in triplicate.

Microarray assay was performed using a service provider (LC Sciences, Houston, Texas). The assay started with a 4-μg total RNA sample. Hybridization was performed overnight on a microfluidic chip consisting of chemically modified nucleotide coding segment complementary to target microRNA (miRBase 13.0) or other control RNA. Fluorescence images were collected using a laser scanner and digitized using Array-Pro image analysis software. Raw data matrix is then subtracted by the background matrix. Data adjustment includes data filtering, Log2 transformation, gene centering, and normalization. A 2-sample t test was conducted for statistical analysis.

Minicircles are the product of site-specific intramolecular recombination between the attB and attP sites driven by bacteriophage ΦC31 integrase (Supplemental Figure I; available at http://circ.ahajournals.org). The DNA fragment containing firefly luciferase and enhanced green fluorescent protein (MC-LG), mir-210 precursor (MC-210), or miR-scramble (MC-Scr) were bluntly ligated between attB and attP sites of p2øC31 minicircle plasmid. The minicircle DNA plasmid (gift from Dr Mark Kay, Stanford) was produced as described previously.⁹ Briefly, Escherichia coli Top10 (Invitrogen) was transformed by parental plasmids. A single colony of the transformants was grown at 37°C overnight. Then 800 mL of bacterial culture was spun down in a centrifuge at 1500 g for 15 minutes at room temperature. The pellet was resuspended with 200 mL of fresh LB broth (pH 7.0) containing 1% L-(+)-arabinose. The resuspended bacteria were incubated at 30°C at 250 rpm for 2 hours. Subsequently, 300 mL of fresh LB broth (pH 8.0) containing 1% L-(+)-arabinose was added and the bacteria were incubated for additional 2 hours at 37°C. Minicircle DNA was isolated using plasmid purification kits from Qiagen (Valencia, Calif).

Adult female FVB mice (10 weeks old) were purchased from Charles River Laboratories (Wilmington, Mass). Ligation of the mid-left anterior descending (LAD) artery was performed by a single experienced microsurgeon (X.W.). Myocardial infarction was confirmed by myocardial blanching and electrocardiographic changes. After 15 minutes, the animals were then injected intramyocardially with 25 μg of minicircles carrying miR-210 precursor (MC-210) or minicircles carrying Scramble (MC-Scr; n=15 per group). Injections were made near the peri-infarct region at 3 different sites with a total volume of 25 μL using a 29-gauge Hamilton syringe. The third group was performed sham, which underwent surgery but not LAD ligation (n=15). Study protocols were approved by the Stanford Animal Research Committee.

Echocardiography was performed before (Day −7) and after (Week 2, Week 4, and Week 8) the LAD ligation. A Siemens-Acuson Sequioa C512 system equipped with a multifrequency (8 to 14 MHZ) 15L8 transducer was used by an investigator (M.H.) blinded to group designation. Analysis of the M-mode images was performed using DicomWorks 1.3.5 (http://dicom.online.fr) analysis software. Left ventricular end-diastolic diameter and end-systolic diameter were measured and used to calculate left ventricular fractional shortening (FS) by the following formula: FS(%)=[(EDD−ESD)/EDD]×100%.

Invasive steady-state hemodynamic measurements were performed (n=5 mice per group for MC-210, MC-Scr, and sham) at Week 4 as described previously.¹⁰ Briefly, after a midline neck incision, a 1.4-Fr conductance catheter (Millar Instruments, Houston, Texas) was introduced into the left ventricle through the right carotid artery. The signals were continuously recorded using a PV conductance system coupled to a PowerLab/4SP analog to digital converter (AD Instruments, Colorado Springs, Colo). Data were analyzed by using a cardiac PV analysis program (PVAN 3.4; Millar Instruments) and Chart/Scope Software (AD Instruments).

Bioluminescence imaging was performed using the IVIS Spectrum system (Caliper Life Sciences). Recipient mice were anesthetized with isoflurane and injected intraperitoneally with d-luciferin (200 mg/kg body weight). Mice were imaged before surgery (baseline scan) and after surgery on Day 3, Week 1, Week 2, Week 4, Week 6, and Week 8. Peak signals from a fixed region of interest were evaluated and signals expressed as photons per second per centimeter square per steradian (p/s/cm²/sr) as described.¹¹

After imaging, mice were euthanized and left ventricular tissue was obtained at 8 weeks after myocardial infarction (MI). Tissue samples were embedded into optimal cutting temperature compound (Miles Scientific). Frozen sections (5 μm thick) were processed for immunostaining. Trichrome stain (Masson; Sigma) was used to determine collagen content of the infarct regions. For each heart, 8 to 10 sections from apex to base (1.2 mm apart) were analyzed. Images were taken for each section to calculate the fibrotic and nonfibrotic areas as well as ventricular and septal wall thickness. Infarct fraction was determined as [fibrotic area/(fibrotic+nonfibrotic area)]×100% as previously described.⁹ To detect microvascular density in the peri-infarct area, a rat anti-CD31 (BD Pharmingen) was used. The number of capillary vessels was counted by a blinded investigator (Z.J.) in 10 randomly selected areas using the picture under a fluorescent microscope, as described previously.⁹ The formalin-fixed and paraffin-embedded explanted hearts were used for terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling (TUNEL) assay. Nuclei undergoing apoptosis were stained with the DeadEnd Fluorometric TUNEL System (Promega) and the TUNEL-positive nuclei were counted and calculated as a percentage of all nuclei according to the manufacturer’s protocol. A subset of the mice (n=3 per group) was euthanized and the peri-infarct region of the left ventricle was dissected under stereoscope for Western blot analysis of Efna3 and Ptp1b.

Transient miR-210 overexpression was obtained by transfection of pSUPER-premiR-210 using Fugene6 (Roche) in HEK-239 cells (ATCC).¹² Transfected cells were harvested in 1 mL per 15 dish of cold lysis buffer supplemented with 5 mmol/L dithiothreitol, 1 mmol/L phenylmethylsulfonyl fluoride, protease inhibitors mixture tablets (Roche), and 100 μm/mL of RNasin Plus (Promega). After 30 minutes at 4°C, samples were precleared by A/G-agarose beads (Santa Cruz) and spun at 4°C for 30 minutes at 20 000 g in a microcentrifuge. Next, the lysates were incubated at 4°C with 2.5 g/plate of antic-myc antibody (9E10; Santa Cruz) for 3 hours and then 50 μL/plate of A/G-agarose beads were added to each sample. After 1 hour, immunocomplexes were washed 2 times with lysis buffer and resuspended in 200 μL of TRIzol (Invitrogen). RNA was purified and specific mRNAs were measured by quantitative real-time polymerase chain reaction (PCR). Average values of 2 genes that were RNA-induced silencing complex (RISC)-associated but not miR-210 targets (B2M and RPL13) were used for normalization.¹³

Statistics were calculated using SPSS 16.0 (SPSS Inc, Chicago, Ill). Descriptive statistics included mean and SE. One-way analysis of variance and 2-way repeated-measures analysis of variance with post hoc testing were used. Student or Welch t test was used depending on Levene test on variance homogeneity. If the probability value of the Levene test was >0.05, Student t test was used or else Welch t test was used. Differences were considered significant at probability values of <0.05.

To identify potential miRNA targets in our study, we first set up miRNA expression profiling experiment. Murine HL-1 cardiomyocytes were subject to hypoxia for 48 hours. Using fluorescence-activated cell sorting, we obtained 2 main populations of cells, apoptotic cells and live cells. We performed miRNA microarray on these 2 populations of cells using the Sanger miRBase Version 13.0 miRNA expression microarrays (LC Sciences, Houston, Texas). We analyzed 679 unique mature miRNAs across biological duplicates of each cell type and found that 7 miRNAs were significantly upregulated and 13 miRNAs were significantly downregulated in live cells compared with apoptotic cells. In particular, miRNA-210 was upregulated approximately 11-fold in live cells (Figure 1A). To confirm the results from miRNA microarray, we selected miR-210 and miR-1187 for real-time PCR analysis. Indeed, miR-210 was significantly upregulated in live cells (Figure 1B), whereas miR-1187 was significantly upregulated in apoptotic cells (Figure 1C). Mir-1187 has been found to be expressed in mouse embryo stem cells by high-throughput pyrosequencing, but its function has not been elucidated yet.¹⁴ We are more interested, however, in miR-210 because it has been shown to be associated with the angiogenesis-associated factor and antiapoptotic gene.^6,15 To further assess whether hypoxia regulates miR-210 expression in cardiomyocytes, we also explored the time-course regulation of miR-210 in HL-1 cells. Induction of miR-210 by hypoxia was discernible at 12 hours of hypoxia, became statistically significant at 24 hours, and increased progressively to approximately 7-fold higher expression by 72 hours (Figure 1D).

To assess angiogenic potential of miR-210, HL-1 cells were transduced by a lentivirus-carrying miR-210 precursor (Pre-210) or by a lentivirus-carrying miR-scramble (Pre-Scr). Under fluorescence microscopy, nearly all the HL-1 cells were green fluorescent protein-positive in both groups, indicating no significant difference in transduction efficiency (Supplemental Figure II). Using real-time PCR analysis, miR-210 expression was 124±15-fold higher in the Pre-210 group compared with the Pre-Scr group. Figure 2A shows that HL-1 transduced with miR-210 could release several angiogenic factors compared with control cells, including Leptin,¹⁶ interleukin-1-α,¹⁷ and tumor necrosis factor-α.¹⁸ In addition, Pre-210 reduced caspase 3/7 activity in HL-1 cells compared with Pre-Scr control under both normoxia (1505 884±84 802 versus 649 933±32 309; P<0.01) and hypoxia (2832 896±97 509 versus 1886 473±48 009; P<0.01) conditions. Conversely, inhibition of miR-210 (anti-210) increased caspase 3/7 activity compared with control anti-Scr in both normoxia and hypoxia conditions (Figure 2B). Moreover, fluorescence-activated cell sorting analysis showed fewer apoptotic cells (22.13±0.48% versus 32.14±1.52%; P<0.05) and more live cells (71.95±1.69% versus 63.39±0.95%; P<0.05) in the Pre-210 group compared with the Pre-Scr group after 48 hours of hypoxia stress (Figure 2C). Taken together, these data demonstrate miR-210 can promote angiogenesis and inhibit apoptosis.

To examine whether miR-210 delivery can improve cardiac function after MI, nonviral minicircles were used to carry the miR-210 expression cassette (MC-210). As novel nonviral vectors, minicircles lack both an origin of replication and the antibiotic selection marker and carry only short bacterial sequences. Their smaller size confers greater transfection efficiency and the lack of bacterial backbone creates less immunogenicity and longer transgene expression.⁹ We first transfected HL-1 cells with different quantities of minicircles carrying Fluc-eGFP (MC-LG) in 6-well plate (Figure 3A). Data showed bioluminescence signals correlated robustly with in vitro Fluc enzyme activity (r²=0.96; Figure 3B). Next, to monitor the duration of transgene expression mediated by MC vector in living animals, a subset of animals with LAD ligation (n=5) were injected with 25 μg of MC-LG into the heart. In vivo bioluminescence imaging indicated that minicircle vector-mediated gene expression was stable for at least 8 weeks in the animal heart (Figure 3C–D). To examine whether MC-210 can improve cardiac function, adult FVB mice underwent LAD ligation and were injected intramyocardially with (1) MC-210; (2) MC-Scr (control); and (3) sham operated animals (n=15 per group). Echocardiography was performed before and after (Week 2, Week 4, and Week 8) the LAD ligation. At baseline, left ventricular FS was comparable in all 3 groups (Figure 4A–B; Supplemental Table I). After LAD ligation, the MC-210 group had significantly higher left ventricular FS compared with the MC-Scr group at Week 4 (28.7±2.4% versus 25.1±1.9%; P<0.05) and Week 8 (27.8±1.9% versus 24.2±2.7%; P<0.05). This finding was further corroborated using invasive PV loops. The PV loop data showed that left ventricular end-diastolic volume and end-systolic volume in the MC-210 group were significantly lower than MC-Scr group, suggesting a more favorable left ventricular remodeling process after miR-210 treatment (Supplemental Table II).

After imaging, animals were euthanized and hearts explanted. Masson trichrome staining showed less infarction size for the MC-210 group compared with the MC-Scr group at Week 8 (Figure 4C), confirming the positive functional data seen in echocardiography. Calculated infarct fractions were significantly smaller in the MC-210 group compared with the MC-Scr group (26.5±2.4% versus 35.4±1.8%; P<0.05). TUNEL staining demonstrated significantly reduced apoptotic cells in the MC-210 group compared with the MC-Scr group (0.13±0.01% versus 0.22±0.04%; P<0.05), whereas few or no TUNEL-positive cells could be detected in the sham group (0.009±0.003%; Figure 4D). Finally, capillary status was evaluated by CD31 immunostaining in the peri-infarct areas (Figure 4E). Capillary density was significantly increased in the MC-210 group compared with the MC-Scr group after MI (362±25 versus 253±37 vessels/mm², P<0.05).

To investigate the mechanism(s) of miR-210-based therapy, we predicted the target genes for miR-210 using both TargetScan and MicroCosm algorithms. Efna3, Dapk1, and Ctgf were predicted to be the putative target genes of miR-210 with high scores. The 3′-UTR segment of these 3 genes containing miR-210 putative binding site (“seed” sequence), which has a crucial role in miRNA:mRNA interaction, were very conserved in different species. Although it was reported that Ptp1b should be 1 of the miR-210 targets, the binding site was not in the 3′-UTR segment.¹³ To confirm that they are indeed the target genes of miR-210, the putative binding site of these target genes (Efna3, Ptp1b, Dapk1, and Ctgf) was amplified by PCR from mouse cDNA and inserted downstream of the luciferase reporter gene in the pGL3 control vector for dual-luciferase assay (Supplemental Figure III). After NIH/3T3 cells were cotransfected with reconstructive vectors and normalizing vector pRL-TK containing Renilla luciferase, the precursor miR-210 mimic (Pre-210) significantly reduced the luciferase activities of the wild-type Efna3, Ptp1b, Dapk1, and Ctgf reporters by 35% to 60% (P<0.05) compared with the miR-scramble (Pre-Scr) control (Figure 5A). In contrast, mutant reporters with 4 nucleotides mismatching noncomplementary seed binding sites were not repressed by miR-210 precursor, confirming that the target site directly mediates repression of the luciferase activity through seed-specific binding. Taken together, these results demonstrate that Efna3, Ptp1b, Dapk1, and Ctgf are the target genes of miR-210.

To further demonstrate that they are direct miR-210 targets, biochemical assay based on the immunoprecipitation of RISC complexes enriched for miR-210 and its targets was used.¹³ To this end, a c-myc-tagged allele of Ago2, a core component of the RISC complex, was used. Easily transfectable HEK-293 cells were cotransfected with expression vectors for miR-210 and c-myc-Ago2, yielding cells enriched of miR-210/c-myc-Ago2-containing RISC complexes as well as miR-210 targets. We then immunoprecipitated c-myc-Ago2 and measured the levels of coimmunoprecipitated mRNAs by quantitative PCR. Background controls were represented by c-myc-immunoprecipitates derived from cells transfected with miR-210, but not c-myc-Ago2, and displayed low to undetectable signals for all the assayed genes (data not shown). Two known miR-210 targets (Efna3 and Ptpt1) were used as positive controls.^12,19 Figure 5B shows that Efna3, Ptp1b, Dapk1, and Ctgf were significantly enriched in immunoprecipitates of miR-210-overexpressing cells compared with cells transfected with a scramble sequence (Pre-Scr). We concluded that Efna3, Ptp1b, Dapk1, and Ctgf are all associated with miR-210 loaded RISC complexes and hence they are the real targets of miR-210.

Although these genes were identified as target genes for miR-210, it is still unknown whether miR-210 could regulate their expression endogenously. Because Efna3 and Ptp1b are involved in vascular remodeling and apoptosis in the heart, respectively, we selected these 2 genes for confirmation in HL-1 cells. HL-1 cells were transduced with Pre-210 to assess whether miR-210 could regulate endogenous Efna3 and Ptp1b. Compared with the control, the level of Efna3 mRNA was significantly downregulated by Pre-210 (Figure 6A) but not Ptp1b. However, we found the level of Ptp1b protein was downregulated on the Western blot (Figure 6B). These data suggest that endogenous Efna3 and Ptp1b are regulated by miR-210 at the level of mRNA and protein, respectively. Our data were also confirmed by immunofluorescence staining (Figure 6C). Furthermore, we also evaluated if miR-210 treatment in the heart would have an effect on Efna3 and Ptp1b. Western blot data from the peri-infarct regions of explanted hearts showed that Efna3 and Ptp1b in the MC-210 group are lower than in the MC-Scr group (Figure 6D). Taken together, these results suggest that overexpression of miR-210 led to downregulation of Efna3 on the mRNA level and downregulation of Ptp1b on the protein level. Efna3 is involved in inhibition of angiogenesis,¹² whereas Ptp1b is involved in induction of apoptosis.²⁰ Therefore, the suppression of these 2 targets by miR-210 delivery may contribute to the improvement of cardiac function after MI.